|
|
Power Analysis for T-Tests: Examples for
Dissertation Students & Researchers
Power Analysis for a Medium Effect Size
For group comparisons using t-tests for independent samples, an absolute mean difference of 0.5 between groups on the variables of interest will be considered meaningful. To achieve a medium effect size (d = .50), a sample of 10 analyzable control subjects in the control and 12 patients with chronic psychotic disorders (total N = 22) provided 30% power to detect such a difference at the 0.05 significance level (critical t (20) = 1.72).
Power Analysis for a Large Effect Size
For group comparisons using t-tests for independent samples, an absolute mean difference of 0.5 between groups on the variables of interest will be considered meaningful. To achieve a large effect size (d = .80), a sample of 10 analyzable subjects in the control and 12 patients with chronic psychotic disorders (total N = 22) provided 56% power to detect such a difference at the 0.05 significance level (critical t (20) = 1.72).
|
|
